Skip to Main Content
 

Global Search Box

 
 
 
 

ETD Abstract Container

Abstract Header

Novel methods for Reduced Energy and Time Consumption for Mobile Devices using Markov Decision Process

Alasmari, Khalid R

Abstract Details

2020, Doctor of Philosophy, University of Toledo, Electrical Engineering.
Computation offloading is a technique for saving energy and time on mobile devices by executing some tasks on the Edge and the Cloud instead of on the mobile devices. Offloading works by adopting MEC technology that has CC capabilities at the edge, which promises to improve the battery life performance. To explore this issue, this dissertation utilizes a MDP-based methodology to make such choices while intelligently optimizing the multiple objectives of latency, energy and security. We provide an analysis of the task offloading problem by adopting different scenarios involving mobile devices users, edge devices, and cloud sites: a small-scale model, a large-scale model, a small-scale-with-preferences model, and a small-scale model in which the preferences change over time. The large-scale model considers a set of mobile device users with computationally intensive components to execute, either on an offloading site or on the mobile device, and determines the quantity of offloading as well as time and energy savings in relation to the number of users and offloading sites available. We formulate divide the problem into two different network topologies: centralized and distributed. The first small-scale model considers three mobile device users with different preferences for where processing takes places, based on their desire to improve latency, power usage, or security, to determine how such preferences affect offloading and energy savings. The second small-scale model considers three mobile device users whose processing preferences change over time, to determine how the MDP interacts with those shifting preferences. The results of the large-scale scenario demonstrate a significant improvement in energy and time performance, one that decreases with an increase in the number of mobile devices. The results of the small-scale models demonstrate that computational offloading algorithms can readily accommodate user processing preferences and that the methodology yields meaningful energy savings regardless of user preferences.
Ahmad Y. Javaid (Committee Chair)
Robert Green (Committee Member)
Mansoor Alam (Committee Member)
Junghwan Kim (Committee Member)
Richard Molyet (Committee Member)
121 p.

Recommended Citations

Citations

  • Alasmari, K. R. (2020). Novel methods for Reduced Energy and Time Consumption for Mobile Devices using Markov Decision Process [Doctoral dissertation, University of Toledo]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1587906971444328

    APA Style (7th edition)

  • Alasmari, Khalid. Novel methods for Reduced Energy and Time Consumption for Mobile Devices using Markov Decision Process. 2020. University of Toledo, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=toledo1587906971444328.

    MLA Style (8th edition)

  • Alasmari, Khalid. "Novel methods for Reduced Energy and Time Consumption for Mobile Devices using Markov Decision Process." Doctoral dissertation, University of Toledo, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1587906971444328

    Chicago Manual of Style (17th edition)